1. Field of the Invention
This invention relates to transmission and reception of signals. More particularly, this invention relates to the encoding and modulation of transmission signals with digital data and the demodulation and decoding of received signals corrupted by channel induced intersymbol interference.
2. Description of Related Art
In the transmission and reception of digital data, channel induced intersymbol interference is the arrival of multiple copies of a transmitted signal that causes the information from one segment of information to corrupt a subsequent segment of information. The multiple copies of the transmitted signals occur generally because of reflections of the transmitted signals within the environment.
The level of the reflection determines the level of degradation of the signal received and the probability of the intersymbol interference causing errors in the reception of the signals. Also, the amount of time required for the reflections to transit from the transmitter to the reflection point and thence to the receiver further determines the quantity of intersymbol interference.
Refer now to FIG. 1 for a discussion of channel induced intersymbol interference in a wireless radio frequency (RF) transmission. In this instance the transmitter 5 is connected to the antenna 10, which radiates an RF signal. Even though the RF signal emanates from the antenna 10 as a contiguous wavefront, it is shown as three separate wavefronts 35, 40, and 50. The wavefront 35 is travels through the atmosphere in a direct path. The wavefront 40 travels through the atmosphere and is reflected from a landmass 30 such as mountains and hills. The reflected wavefront 45 then arrives at the antenna 15. The wavefront 50 travels through the atmosphere and is reflected by buildings 25 of a metropolitan area. The reflected wavefront 55, likewise, then arrives at the antenna 15.
The antenna 15 is connected to the receiver 20. The receiver acquires the transmitted signal and recovers the information modulated within the RF signal. It is apparent that the wavefront 35 travels the most direct route from the transmitting antenna 10 to the receiving antenna 15. The reflected wavefronts 45 and 55, on the other hand, arrive at the antenna 15 at some time later than the wavefront 35. If the amplitude of the reflected wavefronts is larger than the environmental noise, the receiver will acquire the reflected wavefronts 45 and 55 and be indistinguishable from the wavefront 35, except that they will be delayed in time.
Refer now to FIG. 2 for a discussion of the channel induced intersymbol interference within a wireless RF data communication system. The digital data 105 and its synchronizing clock 100 is the input signal to the transmitter 5 of FIG. 1. A common method of modulating the RF signal with the digital data 105 is frequency shift keying (FSK) In frequency shift keying, the voltage level of the digital data representing a digit having a binary 0 causes the RF signal to have a first frequency f0 135 and the digit having a binary 1 causes the RF signal to have a second frequency f1 140.
The transmitter 5 transmits the RF signal 110 from the antenna 10 as described above. The RF signal transits the multiple paths as described with each signal being delayed in time. The direct RF signal λ1 115 arrives at the antenna after a delay δ1 145. The reflected RF signal λ2 120 arrives at the antenna after a delay δ2 150. The reflected RF signal λ3 125 arrives at the antenna after a delay δ3 155.
The direct RF signal λ1 115 and the reflected RF signals λ2 120 and λ3 125 are superimposed upon each other at the antenna and transferred to the receiver 20. The receiver 20 employs superheterodyne techniques to extract the digital data from the differences in frequency that defines the digital data. The data message 130 should change from the binary zero to the binary 1 at the time 170. However, the delayed reflected signals λ2 120 and λ3 125 interfere with the direct received signal 115 and cause an uncertainty 160 of the digital data. The digital data may remain at the binary 0 and is not definitely changed until the time 175. The magnitude of the delayed reflected signals λ2 120 and λ3 125 determines the impact of this intersymbol interference. Similarly, the data message 130 should change from the binary 1 to the binary 0 at the time 10, however, the interference from the delayed reflected signals λ2 120 and λ3 125 might cause the data message 130 to remain at the binary 1 until the time 185. This uncertainty time 165 may cause a misinterpretation of the digital data.
Similar channel induced intersymbol interference can occur in wireless infrared light transmission as illustrated in FIG. 3. In this case the transmitter 205 and the receiver 220 are generally enclosed within a room 200. The transmitter 205 excites the light emitting diode (LED) to emit infrared light. A typical type of transmission is on-off-keying (OOK), where the LED is excited for data having a binary one and turned off for data having a binary zero. The light is transmitted to a photodiode 215. The photodiode 215 is connected to the receiver 220, which recovers the received light signal and demodulates the received signal to extract the transmitted information.
The light signal as transmitted from the LED 210 is a contiguous wavefront, but for illustration it is shown as three separate wavefronts 235, 240, and 250. The wavefront 235 represents the portion of the light signal that travels directly to the photodiode 215. The wavefronts 240 and 250 are transmitted toward the sidewalls of the room 200 and the wavefronts 246 and 255 are reflected to the sidewalls to the photodiode 215. The reflected wavefronts 246 and 255 travel a longer distance through the room 200 to arrive at the photodiode 215 and thus interfere with the wavefront 235 that arrive earlier.
Referring to FIG. 4 for a discussion of the modulation of the light signal as transmitted from the transmitter 205 of FIG. 3. The data clock 300 has the frequency rate that the digital data message is gated within the transmitter 210. The modulation clock (PPM CLK 305) that is used to generate the four-pulse position modulated signal of the transmitted signal 310. Each time slot t1, t2, t3, t4 is divided into four sub-increments s1, s2, s3, s4. One sub-increment s1, s2, s3, or s4 is set to a binary one, in this representation, to represent a two digit binary number. Since only one sub-segment may be occupied for any one time slot t1, t2, t3, t4, the coding can only represent the four possible combinations of the two digit binary numbers. The four-pulse position modulated signal of the data message illustrates the modulation of the four possible binary digit combinations of the dual-bit data and is explained as follows:
Time SlotDual-Bit CodePPM Encodingt1001000t2010100t3100010t4110001
The pulse positioned modulated signal 310 is transmitted by activation of the LED 210. The direct light signal λ1 320 arrives at the photodiode 215 after a delay δ1 340. The reflected light signal λ2 325 arrives at the photodiode 215 after a delay δ2 345. The reflected light signal λ3 330 arrives at the photodiode 215 after a delay δ3 350.
The direct light signal λ1 320 and the reflected light signals λ2 325 and λ3 330 are superimposed upon each other at the photodiode 215 and transferred to the receiver 220. The receiver 220 recovers the pulse positioned data message 335. The pulse positioned data 335 changes from the voltage level of the binary zero to the voltage level of the binary one after the delay δ1 340. However, the delay δ2 345 and δ3 350 of the reflected signals λ2 325 and λ3 330 causes the photodiode to detect the binary one for a time longer than the length of a sub-increment s1, s2, s3, s4 of a time slot t1, t2, t3, or t4. The magnitude of the reflection maybe insufficient to consistently determine this uncertainty time 355. Further, if two symbols such as a 11 followed by a 00 are transmitted the detection of the binary one extending beyond the sub-increment s1, s2, s3, s4 of a time slot t1, t2, t3, or t4 now interferes with a subsequent symbol. The severity of the channel induced intersymbol interference prevents the extension of the bandwidth of communication system beyond the ability of the communication to reliably detect and recover the received data message.
U.S. Pat. No. 6,169,765 (Holcombe) describes an output signal pulse width error correction circuit and method wherein errors in a data signal conforming to a communications protocol having a prescribed duty cycle are corrected by monitoring a duty cycle of the data signal, comparing the duty cycle to a duty cycle reference voltage corresponding to the prescribed duty cycle, and adjusting a pulse width of the data signal to conform to the prescribed duty cycle of the protocol.
U.S. Pat. No. 5,394,410 (Chen) explains a technique for encoding data for serial transmission and the correlative technique for decoding the transmitted data.
U.S. Pat. No. 5,892,796 (Rypinski) illustrates frame format and method for adaptive equalization within an integrated services wireless local area network to prevent fading and intersymbol interference due multiple path radio propagation.
U.S. Pat. No. 6,118,567 (Alameh, et al.) teaches a waveform encoding method and device provide for generating/receiving a power efficient binary intensity modulated optical data signal from a binary source signal which minimizes a time between adjacent pulse transitions and maximizes a pulse peak amplitude for transmission over a low-power wireless infrared link
“Efficient Reconstruction of Sequences,” Levenshtein, et al., IEEE Transactions on Information Theory, January 2001, Volume: 47, Issue: 1, pp. 2-22 introduces and provides solutions for problems of efficient reconstruction of an unknown sequence from its versions distorted by errors of a certain type.
“Spread-Response Precoding for Communication Over Fading Channels,” Wornell, IEEE Transactions on Information Theory, March 1996, Volume: 42, Issue: 2, pp. 488-501 presents “spread-response precoding” to effectively transform an arbitrary Rayleigh fading channel into a nonfading, simple white marginally Gaussian noise channel.
“Linear Complexity Of A Sequence Obtained From A Periodic Sequence By Either Substituting, Inserting, Or Deleting K Symbols Within One Period,” Jiang et al., IEEE Transactions on Information Theory, May 2000, Volume: 46, Issue: 3, pp. 1174-1177 provides a unified derivation of the bounds of the linear complexity for a sequence obtained from a periodic sequence over GF(q) by either substituting, inserting, or deleting k symbols within one period.
“On The Synchronizability And Detectability Of Random PPM Sequences,” Georghiades, IEEE Transactions on Information Theory, January 1989, Volume: 35, Issue: 1, pp. 146-156, investigates the problem of synchronization and detection of random pulse-position modulation (PPM) sequences under the assumption of perfect slot synchronization. Bounds on the symbol error probability and the probability of false synchronization that indicate the existence of a severe performance floor are derived. A way to eliminate the performance floor is suggested by inserting ‘special’ PPM symbols in the random data stream.